Speed-up the gaussian process regression

[vbonvin]

This paper proposes a new algorithm to speed-up the GPR computation:
https://arxiv.org/pdf/1703.09710.pdf

And they have a python code that do that for us:
http://celerite.readthedocs.io/en/latest/

I will try to implement that in PyCS and see how it compares to the current implementation, in terms of speed and efficiency

[drphilmarshall]

Cool! Interested to see how it performs, compared to the older curve shifting approach. It should provide an alternative (Bayesian) route to uncertainties on the time delays and magnifications too.

…

On Thu, Mar 30, 2017 at 1:51 AM, Vivien Bonvin ***@***.***> wrote: This paper proposes a new algorithm to speed-up the GPR computation: https://arxiv.org/pdf/1703.09710.pdf And they have a python code that do that for us: http://celerite.readthedocs.io/en/latest/ I will try to implement that in PyCS and see how it compares to the current implementation, in terms of speed and efficiency — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#18>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AArY95yDD4KLWRU42Ljt3YEwEVr9k8eeks5rq20ggaJpZM4MuDSR> .